About

I’m a Ph.D. student in Institute of Image Processing and Pattern Recognition of Department of Automation at Shanghai Jiao Tong University under the supervision of Prof. Jie YANG.

Before coming to SJTU, I received the BA.Eng. degree in College of Electronic and Information Engineering from Tongji University in 2018. I then received the MA.Eng. degree in Department of Automation from Shanghai Jiao Tong University in 2021.

Research interests

My research interests focus on trustworthy deep learning. Currently, my works are about the robustness of Deep Neural Networks (DNNs) against those pixel-perturbed or distribution-shifted inputs, within the fields of adversarial learning and out-of-distribution detection.

• DNNs trained on clean images show terrible generalization performance on those carefully-designed invisible perturbed images, i.e., adversarial examples, which has raised widespread attention on the adversarial robustness of DNNs.
• DNNs cannot generalize well on data that differs from the training distribution, i.e., In-Distribution (InD). In this regard, it remains a valuable task of detecting whether new samples are from the InD or OoD (Out-of-Distribution) of DNNs in the inference stage.

In addition, I have previously dabbled in the random Fourier features, a sub-field of the kernel methods in machine learning.

Awards

• Outstanding Reviewer for ECCV 2024.